AN EFFICIENT BRAIN TUMOUR DETECTION SYSTEM BASED ON SEGMENTATION TECHNIQUE FOR MRI BRAIN IMAGES

Abstract

In this work, we have proposed an efficient brain tumour detection system based on the segmentation technique for MRI brain images, which can detect tumour and trace the position of tumour in the MRI brain images. To detect the tumour region, segmentation method is used on the basis of threshold and it exactly detects the tumour region with better accuracy. In the proposed work, some pre-processing steps (resizing, binarization and morphological operations) are used for detection of tumour region in MRI brain images. It is being observed that the experimental results of the proposed brain tumour detection system gives better result in comparison to other existing techniques. This study introduces an efficient detection of brain tumour from MRI brain images based on the segmentation of gray matter and white matter. The purpose of this segmentation algorithm is to label tumour objects within the MRI brain image and locate their region based on the white matter. The tumour is extracted from the MRI brain images and its exact position is also determined to classify the volume of tumour. By experimental analysis of various parameters such as PSNR, MSE and volume or percentage of tumour, we have observed that the accuracy of proposed system is more that 90% and it is better than the previous work.